(1) Field of the Invention
This invention generally relates to trajectory control and more specifically to the formation of generic models used to produce guidance parameters that are used in directing a pursuing vehicle to a target vehicle.
(2) Description of the Prior Art
Trajectory control of pursuing vehicles, such as torpedoes, can be classified as "post-launch" or "pre-launch" control. In post-launch control, a pursuing vehicle receives updated guidance information after its launch from a launching vehicle, such as a submarine, until the communications link between the pursuing vehicle and the launching vehicle is no longer intact. U.S. Pat. No. 5,319,556 (1994) to Bessacini discloses one embodiment of a post-launch control system with an adaptive trajectory apparatus and method for providing, at and after the launch, vehicle control commands to steer a torpedo (a pursuing vehicle) from a submarine (a launching vehicle) toward a contact (a target vehicle). The development of the commands depends, in part, on the information received from a generic model of the torpedo that is launched.
In a pre-launch control system, the pursuing vehicle, or torpedo, receives all the guidance parameters prior to launch. The control system responds to estimates of current target vehicle state and classification to establish target vehicle operating characteristics in order to project an anticipated target vehicle trajectory. A representation of a pursuing vehicle characteristic trajectory derived from a corresponding generic model of the pursuing vehicle provides a projected pursuing vehicle trajectory based upon initially provided parameters. Iterative processing of the functional forms of these two trajectories, starting with the initially provided parameters, provides successive operating parameter solutions that converge to generate the guidance parameters that are transferred to the pursuing vehicle immediately prior to launch. Since the computation of these guidance solution parameters must be performed every update cycle of the control system, the iterative processing must converge to the guidance solution within each update cycle. The development of these parameters, therefore, is dependent upon the information received from a generic model of that pursuing vehicle.
Both post-launch and pre-launch systems therefore depend upon information in a generic model of the pursuing vehicle. Consequently, to a significant degree the accuracy of the guidance commands or parameters supplied to the pursuing vehicle is dependent upon the accuracy with which the information in the generic model describes the actual trajectory of the pursuing vehicle.
A generic model must, as known, take into account the physical characteristics of the pursuing vehicle under a variety of kinematic states. One approach has been to define the operations of the pursuing vehicle through a set of one or more ballistic constants. For example, U.S. Pat. No. 3,566,743 (1971) to Frohock discloses a kinematic device for fire control against terrestrial targets with a single rate sensor. A ballistic calculator in this system, for example, provides appropriate ballistic values that correspond to the characteristics of a round being fired to develop a ballistic correction that can account for the difference between ballistic trajectory and the line of sight. This is a single plane vertical correction and involves only one ballistic constant. U.S. Pat. No. 5,379,966 (1995) to Simeone et al. discloses a missile guidance system for kinematic states that produces initial tracking information based upon a model. The system then reverts to sensed position information for the projected missile trajectory. Both of these systems rely upon models for anticipating the trajectory of a pursuing vehicle.
Other systems also rely on a vehicle model. U.S. Pat. No. 5,071,087 (1991) to Gray discloses a method for guiding an in-flight vehicle to a desired flight path. U.S. Pat. No. 5,082,200 (1992) to Gray discloses a method for guiding an in-flight vehicle toward a target. U.S. Pat. No. 5,435,503 (1995) to Johnson et al. discloses a real time missile guidance system. Each of these systems relies upon some type of pursuing vehicle model to generate an initial set of flight conditions or to assist in the tracking of a particular vehicle.
Initial approaches for producing generic models for torpedoes as pursuing vehicles involved the in-water testing of actual torpedoes. In essence, torpedoes were launched with known guidance parameters or presets and tracked. The measured trajectory information from multiple tests for a given set of presets was combined to produce an average trajectory that, in turn, yielded a basic set of ballistic constants for the generic model for that set of presets. To encompass the spectrum of possible geometries (tactical situations) and presets, a large number of ballistic constants need to be determined requiring an enormous number of runs to be made. This approach is extremely undesirable. The most important drawback is that the in-water runs are extremely costly and time-consuming. Thus, the number of runs required to ensure robust generic model operation for underwater trajectory systems cannot be made. In addition, any modification to torpedoes requires that all the runs be remade thereby resulting in excessive cost and unacceptable time delays.
More recently there has been developed a six-degree of freedom model simulator that, for a given set of input conditions and characteristics, simulates the track of a torpedo or similar pursuing vehicle for any specified run. Data from each run and from each group of runs for a given set of input conditions and characteristics are then analyzed for determining the ballistic constants based upon average performance. This is a high fidelity simulator that has essentially eliminated the need for actually firing torpedoes. However, for recent torpedo applications the ballistic constant matrices have become quite extensive (i.e., thousands of entries) and require hundreds of thousands of runs to be generated. While the generation of run data can be done must faster using this high fidelity model, the task of analyzing, extracting, and averaging the ballistic constants is still done on a run-by-run basis. This approach is tedious and time-consuming and restricts the number of runs that can be processed. A partitioning of the operation of the vehicle into segments or phases allows for the characterization of the operational features important to the generic model in the trajectory control system. Parameters referred to as ballistic parameters that are sets of constants are determined for each of the phases. The phases and associated ballistic parameters allow for the automatic determination of the sets of ballistic constants for each phase in an efficient manner. Consequently, none of the current trajectory model systems, including the aforementioned models disclosed in the above-identified Untied States Patent Letters, incorporate any mechanism for the automatic extraction of sets of ballistic constants from six degree of freedom simulations.